Titration without separation, e.g. quantification of a target species in living cells, is a challenge of analytical chemistry. We perform the selective detection of a target using the kinetics involved in a photochemical process and develop a correlation method that we illustrate by the titration of a fluorescent photoswitcher and the target of a photoswitching sensor. Correlating an input time series and a well-chosen weighting function associated with a variable characteristic time yields a spectrum of characteristic times. The upper integration limit of the correlation output can be chosen to match the argument of an extremum of the spectrum with a characteristic time of the input time series in order to quantify the target. A similar procedure is followed to optimize the signal-to-noise ratio. Selectivity and signal-tonoise ratio associated with 15 weighting functions are theoretically predicted. The results are applied to the titration of the reversibly photoswitchable fluorescent protein Dronpa-2 and the titration of calcium using a reversibly photoswitchable fluorescent sensor. The performance of the correlation method is favorably compared to the one of other dynamic contrast protocols.