Semapondo and Proof in Land Review Cases

The Supreme Court decision in Semapondo v Rwasimpenzi et al is a useful Rwandan precedent on review due to injustice and proof of land ownership. Its full citation is Semapondo v Rwasimpenzi et al, Supreme Court, [2023] 1 RLR 1; RS/INJUST/RAD 00002/2020/SC, Supreme Court, 10 December 2021. The case was decided by Ntezilyayo, P.J., with Nyirinkwaya, Cyanzayire, Hitiyaremye and Kalimunda, J. Although the dispute began as a disagreement over one plot in Kigali, the judgment speaks to a wider problem in land litigation: what kind of evidence is enough to disturb a registered land position after ordinary appeals have already run their course? The answer is not as simple as saying that official documents always win or that witness testimony always fails. The decision is more careful than that. It suggests that a court will look for a convincing chain between the alleged payment, the land claimed, the seller, the plot number, and the public land record.

The case appears especially important because review due to injustice is an exceptional remedy. A party seeking that remedy is not merely asking the Supreme Court to read the file again with fresh sympathy. The party is asking the Court to accept that the earlier judgment produced a legally recognisable injustice. That is a serious claim, and it needs serious proof. In Semapondo, the applicant had evidence that he paid money to Association Goboka. What he struggled to prove was that the payment related to the specific disputed plot, later registered as UPI 1/02/09/02/1418. That distinction may look small in ordinary conversation, but in land law it is often everything.

Semapondo Charles said that on 19 November 2007 he deposited 4,900,000 Rwandan francs into the account of Association Goboka at Banque Populaire du Rwanda. He argued that the payment was for a plot then known as number 12438, later identified as number 1418, situated in Nyirabwana Village, Kibagabaga Cell, Kimironko Sector, Gasabo District, Kigali City. When he tried to register the land, he found that the records of the Rwanda Land Management and Use Authority showed the same plot as belonging to Rwasimpenzi Wellars. That discovery pushed the dispute into court. Semapondo asked for documents from Goboka Cooperative, cancellation of Rwasimpenzi's long term lease, registration in his own name, and damages.

The Intermediate Court of Gasabo accepted Semapondo's claim. It annulled the long term lease held by Rwasimpenzi, ordered registration in Semapondo's name, and awarded damages against Goboka Cooperative. Rwasimpenzi appealed to the High Court. He argued that the claim should not have been admitted because the property had already been the subject of another final case, and he also complained that the trial court had ignored evidence from the defence. The High Court accepted the objection and quashed the Intermediate Court judgment.

Semapondo then appealed. After judicial reforms, the matter reached the Court of Appeal. That Court disagreed with the High Court on admissibility and held that Semapondo's claim should have been heard. Yet the Court of Appeal did not agree with him on ownership. On the merits, it found that plot number 1418 belonged to Rwasimpenzi, maintained the long term lease between Rwasimpenzi and the Government of Rwanda, and ordered Semapondo to pay procedural and counsel fees. Semapondo later sought review due to injustice, and the President of the Supreme Court approved that the judgment be reviewed.

The Supreme Court had to decide whether the Court of Appeal judgment was affected by injustice because it allegedly disregarded evidence showing that Semapondo was the real buyer. The central question was not whether Semapondo had ever paid money to Goboka. The Court appears to have accepted that the payment was real. The harder question was whether the payment was proved to relate to plot number 12438, later registered as plot number 1418. That meant the Court had to compare the strength of a general bank deposit slip and witness statements against documents that identified the plot with more precision.

Semapondo relied mainly on the bank deposit slip, witnesses who claimed to know that the plot had been sold to him, and the argument that plot numbers can change over time. Rwasimpenzi relied on a broader documentary trail. His evidence included an invoice identifying the plot, a bank deposit slip connected to that plot, a plot allocation certificate, property titles, land titles issued by the land authority, the long term lease agreement, and a previous judgment from Kacyiru Primary Court indicating that no double agreement existed for the disputed plot. The contrast was sharp. One side had proof of payment, but weak identification of the land. The other side had proof that appeared to tie payment, allocation, title, and registration to the same parcel.

The Supreme Court grounded its reasoning in the ordinary burden of proof. It referred to article 12 of Law No. 22/2018 of 29 April 2018 relating to civil, commercial, labour and administrative procedure, under which a claimant must prove the claim, and if the claimant fails, the respondent wins. It also relied on article 3 of Law No. 15/2004 of 12 June 2004 relating to evidence and its production in courts, which places on each party the burden of proving the facts that party alleges. Those provisions may sound familiar, even routine. In this case, however, they carried real weight. The applicant was not relieved of proof merely because review had been allowed to proceed.

The rule emerging from the decision may be stated in this way. In a review due to injustice involving land ownership, the applicant must prove the alleged injustice through evidence that reliably identifies the disputed land and connects the applicant's alleged transaction to that land. A bank deposit slip, even if genuine, is likely to be insufficient where it does not mention the plot number, location, area, or other identifying details, especially if the opposing party produces official documents and transaction records pointing to the same land. The rule does not make official registration untouchable. Rather, it requires anyone challenging that registration to bring evidence with enough specificity to justify the challenge.

The Court first rejected the argument that the review should not be heard on the merits. It relied on Busoro Gervais v Busoro Mugunga Desire et al, RS/INJUST/RC 00022/2018/SC, Supreme Court, 28 June 2019. In that earlier case, the Supreme Court had held that once the President of the Supreme Court approves review due to injustice, the Court proceeds to examine the case on the merits, subject only to objections of public order. Semapondo therefore confirms the practical effect of Busoro. Approval of review opens the door to a merits hearing, but it does not predetermine the result.

On the evidence, the Supreme Court found that Semapondo had not proved that the Court of Appeal judgment caused the injustice alleged. The bank slip proved a payment to Goboka, but it did not identify the disputed plot. The witness evidence did not close that gap. One witness had even signed a later document confirming Rwasimpenzi as the buyer of the disputed plot, which weakened the reliability of the testimony relied on by Semapondo. By contrast, Rwasimpenzi's documents identified the disputed land by number, location, area, allocation, title, and lease. The seller's confirmation also supported his version. The Court held that UPI 1/02/09/02/1418 belonged to Rwasimpenzi Wellars and sustained the long term lease agreement between Rwasimpenzi and the Government of Rwanda.

The financial order also mattered. The Supreme Court ordered Semapondo to pay Rwasimpenzi and Goboka Cooperative 800,000 Rwandan francs each for counsel and procedural fees, and to pay the Rwanda Land Management and Use Authority 100,000 Rwandan francs. That order shows that the Court did not view the review as a harmless procedural step. Once a party brings another party back to court through review and fails, costs may follow.

The case matters because it protects review due to injustice from becoming an informal second appeal. It is easy to understand why a losing party in a land dispute may feel that something has gone wrong, especially where land has personal, family, and economic meaning. Still, a feeling of unfairness is not the same as proof of injustice. Semapondo teaches that the Supreme Court may reconsider the case once review is approved, but the applicant must still show that the earlier judgment was wrong in a legally relevant way. The Court did not punish Semapondo for relying on a bank slip. It simply held that this particular slip did not identify this particular plot.

The decision is also useful for land administration. Rwanda's land system depends on precise identification of land. Plot numbers, locations, areas, allocation documents, titles, and lease agreements are not technical decorations. They are the tools through which courts and land authorities prevent the same land from being claimed by several people on vague evidence. A receipt that says money was paid may be important, but it must still answer the next question: paid for what land? Without that link, the court may hesitate to cancel a registered title.

At the same time, the judgment should not be read in an overly formal way. It does not appear to say that oral testimony has no place in land disputes. It considered the witnesses. It also considered the argument that plot numbers can change. The problem was that the oral evidence did not fit comfortably with the written and official record. That is a more balanced lesson. Courts may listen to witnesses and examine factual history, but where witness accounts are inconsistent or general, and where official documents tell a more coherent story, the specific documents are likely to carry greater force.

For buyers, the practical lesson is simple but often ignored. Do not rely only on proof that money moved. A buyer should make sure that the purchase documents identify the plot number, location, boundaries, area, price, seller, buyer, and date. If the land is later renumbered, there should be a traceable document explaining the change. Registration should be pursued promptly, and communication with the seller, cooperative, local authorities, and land authority should be kept. In real life, people sometimes trust a receipt because the seller is known in the community or because the transaction happened through a cooperative. Semapondo shows why that trust can become fragile in court.

For lawyers, the case is a warning about how to build a review due to injustice application. It is not enough to say that the Court of Appeal ignored evidence. Counsel must show that the evidence was strong enough to change the legal result. The better approach is to map every document onto the disputed plot and explain why the rival registration cannot stand. If the applicant relies on a payment slip, the lawyer should be ready to show how that payment connects to the land through a contract, allocation record, invoice, correspondence, witness evidence, or official registration history. The missing link in Semapondo was exactly that connection.

Semapondo also fits with later and related Rwandan cases on review due to injustice. Its reference to Busoro confirms the procedural rule that approved review is heard on the merits. Its relationship with cases such as Sonarwa Life Assurance Company Ltd v Rugina may suggest a wider pattern: review due to injustice is available, but it is not a way to escape the normal discipline of proof, time limits, and procedural diligence. That approach is sensible. It leaves space for correction of real injustice, while avoiding endless reopening of disputes simply because a party remains dissatisfied.

Semapondo v Rwasimpenzi et al is not important merely because one party won and another lost. Its deeper value lies in the way it connects review, evidence, and land registration. A general payment document may be genuine and still fail to prove ownership of a particular plot. Witness testimony may support a claim, but it must be consistent, specific, and compatible with the documentary record. Registered land documents, allocation records, invoices, titles, and long term leases are likely to carry strong evidential weight where they point in the same direction. The decision leaves a useful message for litigants: review due to injustice is not won by suspicion, sympathy, or broad narrative. It is won, if at all, through clear proof that connects the alleged wrong to the legal right claimed.

Source note. This article is based on Semapondo v Rwasimpenzi et al, Supreme Court, [2023] 1 Rwanda Law Reports 1; RS/INJUST/RAD 00002/2020/SC, judgment of 10 December 2021, with reference to the statutory rules on proof in Law No. 22/2018 of 29 April 2018 relating to civil, commercial, labour and administrative procedure and Law No. 15/2004 of 12 June 2004 relating to evidence and its production. It is prepared for public legal education only and should not be treated as legal advice for any specific land dispute.

Suggested citation

Ronald Serwanga, “Semapondo and Proof in Land Review Cases” East Africa Legal Insight (2 August 2026).