Two step surgery processes during administration and withdrawal of conventional implants and also the initial burst effect leading to poor drug loading have been considered as major complaints associated with in situ drug implants. The objective of this research was to develop in situ forming implant with highly water soluble drug Metoprolol Tartrate and biodegradable polymer DL-PLA to minimize these problems. Formulations were attempted to be developed with the incorporation of six different hydrophobic excipients namely, Arachis oil, Glyceryl monostearte, Stearic acid, Magnesium stearate, Cetostearyl alcohol and Stearyl alcohol. Effects of drug: polymer ratio and presence of hydrophobic excipients in formulations were observed on extent of drug burst, implant loading efficiency, duration of drug release and possible drug release mechanism. Loading efficiency of 5% theoretically drug loaded implant was found to be more than that with 10%. The lower drug loaded formulation more efficiently controlled the extent of drug burst while also extending the duration of drug release for a longer period of time. While addition of Arachis oil, Glycerol monostearte, Stearic acid, Magnesium stearate in 5% theoretically drug loaded formulations improved drug loading efficiency as compared to implants without excipients, Cetostearyl alcohol and Stearyl alcohol did not exert any such effects. Extent of burst was reduced as well as duration of drug release extended for Arachis oil, Glycerol monostearte and Stearic acid incorporated implants. Even though improved loading efficiency was observed with Magnesium stearate, an insignificant raise in burst and no change in duration of release were also observed. While with Cetostearyl alcohol and Stearyl alcohol the duration of release remained the same, the former increased the burst and the later reduced the same as compared to drug only implants. The release profile exhibited biphasic behavior in all the cases, an initial burst release phase followed by steady state drug release. The kinetics was evaluated for the both phases by fitting the data in four different kinetic models, namely, Zero order, First order, Higuchi and Korsmeyer-Peppas. While the burst phase fitted best with Zero order kinetics the steady-state phase complied more with Higuchi model.