The form in which the national product affects the level of unemployment has become one of the most relevant macroeconomic subjects. A simple form of measuring the relation between the increase (decrease) of the national product and the decrease (increase) of the unemployment rate is explained by the so called “Okun Law” according to which it is not expected, except under very special conditions, that each point of increase of the product results in a point of increase of the employment and, consequently, a fall of one point in the unemployment rate. The author shows a relation between growth and unemployment in the following equation ut – ut-1 = – p(gt – ct), where ut is the unemployment rate, gt the percentage growth of the product and ct represents the percentage growth of the product that it is needed to maintain the unemployment level of the previous period. One of objectives of this study is to estimate ct. In the economic formulation of this equation, this work takes into consideration that, if deterministic trends included within the explicative variables, the resulting model will necessarily be very restrictive in the sense that, any impulse no matter how intense it was, will not have effects in the long run, because everything will be captured by the given trend. The necessary flexibility is introduced by allowing the trend to evolve in time as a stochastic process, which in turn is equivalent to establishing the “stylized facts” associated with these time series. The approach is therefore on the introduction of a stochastic trend in the regression model, seeking to take into account the structural changes that occurred during the most recent years in the Argentinian economy. From the statistical point of view, the key way for modeling a stochastic trend is the state space form. This methodology allows the unknown parameters to be estimated via the prediction error decomposition and the forecastings to be computed extending the Kalman filter. Moreover, a smoothing algorithm based on the Kalman approach can be used to provide an optimal estimator of the trend at each point of the sample period.