Thirty-nine linear regression and time series models were built and calibrated for influent temperature (Ti) estimation at the primary aerated facultative lagoon in a municipal wastewater treatment plant. The models were based on mean daily ambient air temperature (Ta) and/or daily rainfall (P), and—optionally—wastewater temperature autoregression. The best fits were achieved with some time series models involving Ta and P, and Ti autoregression. The best-fit model was able to estimate influent temperature with a root-mean-square-error of 0.5 °C, and an R2 of 0.925, for the calibration period of 10.5 months. In addition, a dynamic lagoon-temperature (Tw) model from the literature was modified in its terms of solar radiation and aeration latent heat, and applied to the primary lagoon. The model was fed with the estimated influent temperature, and five model parameters were identified by calibration against 10.5-month Tw data. Dynamic lagoon-temperature estimation results were comparable to or better than other results of long-term simulations found in the literature. Sensitivity analyses were run on both models. Further validation with independent sets of data is needed for verification of the predictive capability of the models.