Many studies have addressed the issues of the price relationships between spatially separated commodity markets. Such studies, which have policy relevance, provide information on the structure and the performance of the markets. Most of these studies, however, have been criticized for not considering the transfer costs and the seasonality in their analyses. In this study, taking two spatially separated markets in Nepal, we estimate and test for the spatial pricing efficiency in tomato markets. Specifically, we estimate the speed of price adjustment in the local market in response to the price shocks in the central market, and examine the efficiency of spatial arbitrage between these two markets. The data we use is unique in that the data was collected from each market every three days. Moreover, the day of data collection was synchronized across the markets. The model we specified is an extension of Ravallion's dynamic model including both transfer costs and seasonality. The results showed that about 12% of the price shocks in the central market was transmitted to the local market on the same day, while 67 % of it was transmitted in a period of three days. However, the effect of transfer costs on the price spread between the markets was negligible. The results also showed an incomplete spatial arbitrage indicating the inefficiency of local market. The results also indicated that estimation without seasonality posed a serious autocorrelation problem. This problem was improved when the seasonality component was included in the analysis. The effect of seasonality was found to have a significant effect on the spatial price efficiency of the markets.