Although robot search and monitoring are two problems which are normally addressed separately, this work conceives the idea that search and monitoring are both required in realistic applications. A problem of simultaneous search and monitoring (SSM) is studied, which innovatively combines two problems in a synergistic perspective. The single pursuer SSM of randomly moving or evasive targets are studied first, and are extended to the cases with multiple pursuers. The precise mathematical frameworks for this work are POMDP, POSG and Dec-POMDP. They are all intractable and non-scalable. Different approaches are taken in each scenario, to reduce computation cost and achieve online and distributed planning, without significantly undermining the performance. For the single pursuer SSM of randomly moving targets, a novel policy reconstruction method is combined with a heuristic branching rule, to generate a heuristic reactive policy. For the single pursuer SSM of evasive targets, an assumption is made and justified, which simplifies the search evasion game to a dynamic guaranteed search problem. For the multiple-pursuer SSM of randomly moving targets, the partial open-loop feedback control method is originally applied to achieve the cooperation implicitly. For the multiple-pursuer SSM of evasive targets, the assumption made in the single pursuer case also simplifies the cooperative search evasion game to a cooperative dynamic guaranteed search problem. In moderate scenarios, the proposed methods show better performance than baseline methods, and can have practical computation efficiency. The extreme scenarios when SSM does not work are also studied.