Antibiotics are highly selective drugs with a great power against micro organisms. They´re quite efficient medicine and usualy do not cause relevant side effects. These drugs can be sometimes considered to act as biological filters selecting resistant populations, and consenquently causing drug innefficiency. Investigatios searching for less empirical approaches of treatment, clearly are of crucial importance. The main goal is to achieve cure, without cause side effects, in the shortest period of time. Here, mathematical models describing seric and tissue kinetics, with single and multiple dosage regimen, are proposed. The models´parameters are estimated by fitting curves to in vivo data, using a non-linear method of optimazation, Hooke and Jeeves. Simulations of these kinetics are presented and discussed. Dosage regiments are designed, in order to follow a pre-determined kinetic behavior. Micro organisms dynamics under antibiotics action are modeled and simulated.