It depends what sort of catalyst you are working with homo or heterogenous catalysts.

With heterogenous catalysts (e.g. Pd/C for hydrogenation), the active sites could be blocked or so called "poisoned" leading to reduced reactivity.

Reused catalyst is poisoned eg PD/ C is easily poisoned.And other catalyst have problems that there surface active sites are blocked or reduced. As surface area plays a major role in nanocatalysis.

Avtar;

Eugene is correct, there are always some reactions that lead to coking, poisoning and general fouling that lead to activity losses in the catalyst. For heterogeneous transition metal type catalyst you may be able to regenerate the active sites through thermal, chemical, or mechanical means. See references below.

www.mdpi.com/2073-4344/5/1/145/pdf

http://www.documentation.emersonprocess.com/groups/publicreadonly/documents/webpage/ad129_ref.hcsp

The case of heterogeneous catalysts has already been taken care of in the above discussions. I would like to cite the example of zeolite catalysts, that are used in  various reactions. In such cases, catalyst deactivation can happen due to coking of the catalyst, giving rise to killing of the active sites. Formation of oligomeric by-products can also choke the pores, which can also cause deactivation.   

In case of homogeneous catalysts, there can be a few reasons for the reduction in catalytic activity.  Here, the catalyst, in general, remain in the distillation residue. 

Some of the reasons could be, (a) decomposition of the catalyst (sulphonic acid catalysts), (b) presence of polymeric material formed as by-product (for example, in case of trans-esterification of acrylic acid esters), which can camouflage the catalyst  (c) relatively large amount of residue formation, that can effectively reduce the catalyst concentration in the reaction mixture.  

General reasons for catalyst deactivation on heterogeneous catalyst like Zeolite or metal doped zeolite are 

1. In case of  Zeolites, they have Bronsted and lewis acidic site. Bronsted acidic sites are the centers for coke formation so if the catalyst has very high Bronsted acidic sites it may give high yield and will also deactivate faster due to coke.

Lewis acidic sites are less prone to coking than Bronsted acidic sites.

Coke contains polyaromatics which mask the surface of catalyst.

2. Due to coke formation, the pores and channels of zeolite will block over a period of time reducing conversion of reactant and ultimately completely deactivation of catalyst.

3. If the catalyst is regenerated by burning off the coke still the destroyed Bronsted and Lewis are never regenerated. So every regeneration of used catalyst will have less acidic sites hence selectivity decreases.

4. After reaction the crystallinity of the catalyst will also reduce so less crystalline catalyst less selectivity and conversion. 

5. Coke formation on zeolite depends on Si/ Al ratio , less Si/ Al ratio more acidic zeolite and hence will also form more coke.