The primary objective of this paper is to test the hypothesis of efficiency wage in the context of Tanzania labour market. The test is facilitated via estimating the correlation between firm level productivity and firm level weighted average wage in Tanzania manufacturing enterprises. The study uses panel dimension of the data to estimate Cobb Douglas’ production function that controls for time invariant characteristics. It also estimates translog production functions to allow for factor substitution. Estimates based on Cobb Douglas production function suggests that controlling for firm fixed effects, one percent increase in real wage results into 0.28 percent increase in productivity. The estimates are stable even when translogs are estimated, though they are reduced up to 0.2. Such findings are taken as evidence of the existence of efficiency wage in Tanzania’s labour market, as wage level can account for about twenty percent of observed productivity in a firm. The paper concludes that firms can use efficiency wage as a policy to induce increased productivity. Partly, this may be due to the fact that a wage premium above the market tends to discourage worker shirking, reduce worker turnover, and mitigate worker adverse selection behaviour. Finally, it is evident that efficiency wages are highly correlated with unobserved firm specific characteristics, suggesting that firms with some characteristics self-select into payment schemes that have efficiency wage aspects.