In this study, we examine the regional income distribution in Peru from 1795 to 2017. To achieve this goal, we reconstructed long-term regional GDP and population series for Peru’s 24 departments. These series allowed us to analyze regional income inequality through dimensions such as inequality, modality, mobility, agglomeration, and convergence. The results indicate a persistent increase in regional inequality in Peru from the second half of the 19th century to the first half of the 20th century. The Gini coefficient, which measures regional inequality, shows a value of 0.2613 for 1795 and 0.3626 for 2017, with the highest value of 0.4283 recorded in 1934. The regional income distribution is bimodal, with no mobility between the extremes. For instance, the probability that a department poor in 1795 remains poor in 2017 is 94%, while the probability of a rich region remaining rich is 95%. However, significant mobility is observed among departments occupying the middle of the distribution. Additionally, the beta convergence rate from 1795 to 2017 was 1.62%, compared to 1.30% in the 19th century and 1.05% in the 20th century. Using Quantile Regressions (QR), we found that the convergence speed for the entire analysis period ranges from 0.5% to 3.22%, depending on the quantile analyzed. In contrast, using Markov-Switching models (MS), we found a convergence speed exceeding 10%, contrary to previous empirical findings. Finally, the impact of geographic variables on convergence speed varies depending on the statistical method used and the period analyzed.