Sesame paste (tahin) is produced by milling hulled, roasted, sesame seeds. In this study, a hot-air roasting process for the production of sesame paste was optimised by response surface methodology (RSM) over a range of air temperatures (120-180 degrees C for various times (30-60 min). The colour parameters (L, a and b values), browning index (131), hardness, fracturability and moisture content of the seeds were used as response parameters to develop predictive models and optimise the roasting process. Increases in roasting temperature and time caused increases in the a and b values and in the BI. The hardness and fracturability of seeds also decreased with increasing roasting temperature and time. The quadratic and linear models developed by RSM adequately described the changes in the colour values and textural parameters, respectively. The result of RSM analysis showed that all colour parameters and textural parameters should be used to monitor the roasting of sesame seeds in a hot-air roaster. To obtain the desired colour and texture, the optimum roasting range for production of sesame paste was determined as 155-170 degrees C for 40-60 min. (c) 2006 Society of Chemical Industry.