In this paper a depletion model with two different approaches (stochastic and deterministic) was analyzed, in order to compare the observation error hypothesis in the catch-per-unit effort (CPUE) data, and the process error in the model, as well as its effect on the recruitment and catchability estimates in the jumbo squid fishery from the Gulf of California, Mexico. Results showed an underestimation of the catchability (q) when the deterministic aproach was used. The observation error showed a bias in q (+31%) higher than in recruitment (+8%), while a Monte Carlo simulation estimated a negative bias for q, and a positive bias in recruitment. The computation of the expected value of catchability ( E(ˆqCPUE/Ci)) was 2.8 × 10–4, while the confidence intervals with the deterministic model were 3.3 × 10–4 < q < 4.1 × 10–4, showing that the estimate of E(ˆqCPUE/Ci) was not within these confidence intervals. E(ˆqCPUE/Ci) fitted the model to the data, describing the trend of the CPUE index in time, and its implications in the management of the jumbo squid fishery.