Abstract Quantification of bacteria being grazed by microzooplankton is gaining importance since they serve as energy subsidies for higher trophic levels which consequently influence fish production. Hence, grazing pressure on viable and non-viable fraction of free and particle-associated bacteria in a tropical estuary controlled mainly by protist grazers was estimated using the seawater dilution technique. In vitro incubations over a period of 42 h showed that at the end of 24 h, growth coefficient (k) of particle-associated bacteria was 9 times higher at 0.546 than that of free forms. Further, ‘k’ value of viable cells on particles was double that of free forms at 0.016 and 0.007, respectively. While bacteria associated with particles were grazed (coefficient of removal (g) = 0.564), the free forms were relatively less grazed indicating that particle-associated bacteria were exposed to grazers in these waters. Among the viable and non-viable forms, ‘g’ of non-viable fraction (particle-associated bacteria = 0.615, Free = 0.0086) was much greater than the viable fraction (particle-associated bacteria = 0.056, Free = 0.068). Thus, grazing on viable cells was relatively low in both the free and attached states. These observations suggest that non-viable forms of particle-associated bacteria were more prone to grazing and were weeded out leaving the viable cells to replenish the bacterial standing stock. Particle colonization could thus be a temporary refuge for the “persistent variants” where the viable fraction multiply and release their progeny.