Background We evaluated if the survival benefit of adding rituximab to standard chemotherapy for non-Hodgkin lymphoma (NHL) observed in clinical trials has been experienced by an Australian NHL patient population. Methods NHL cases diagnosed in 1985-2004 in New South Wales (NSW) were followed-up to the end of 2004. Rituximab prescription data were obtained from Medicare Australia. Using a Poisson regression model adjusted for age group, sex, NHL subtype and time period (1990-1994, 1995-1999 and 2000-2004), we estimated excess risk of death after a diagnosis of NHL. To give context to the survival trend, trends in incidence and mortality were also estimated. Results Compared with 1990-1994, after adjusting for age, sex and NHL subtype the relative excess risk of death was significantly lower (p < 0.0001) in 1995-1999 (0.89) and 2000-2004 (0.74). A sharp fall in mortality was observed from 2000 to 2004 (annual percentage change (APC) = -4.7, p = 0.009), while a small but significant rise in incidence was seen from 1990 to 2004 (APC = 0.5, p = 0.01). The number of times rituximab was dispensed in NSW increased rapidly from 1274 in 1999 to 9250 in 2004. Conclusion It is likely that some benefit of adding rituximab to the standard chemotherapy for NHL has been experienced at the population level.