The split between "typical" (first-generation) and "atypical" (second-generation) antipsychotics is one of the most influential — and most overstated — distinctions in modern psychiatry. For most of the 1990s and 2000s, the new atypicals were marketed as a categorical advance: similar efficacy, fewer movement side effects, less stigma. By the late 2000s, two of the largest pragmatic trials in the history of psychiatry — CATIE in the US and CUtLASS in the UK — had complicated the story considerably. The honest summary today is that both classes have real strengths and real costs, and the choice between them belongs in a careful conversation with a prescriber, not in a marketing pitch.
Atypicals generally cause fewer movement disorders but more metabolic problems than typicals; on most other measures, the two classes are more similar than the original marketing implied.
Defining the classes
First-generation (typical) antipsychotics were developed beginning in the 1950s. Examples include chlorpromazine, haloperidol, fluphenazine, perphenazine, and trifluoperazine. They are characterised by strong dopamine D2 blockade and relatively few off-target effects.
Second-generation (atypical) antipsychotics began with clozapine's reintroduction in the 1980s and expanded through the 1990s. Examples include clozapine, risperidone, olanzapine, quetiapine, aripiprazole, ziprasidone, and many newer agents. They tend to have additional 5-HT2A serotonin blockade and other receptor activity beyond dopamine.
The original promise — and where it broke down
Atypicals were marketed as having three main advantages: equal efficacy, fewer extrapyramidal side effects (EPS), and broader benefit (including for negative symptoms and cognition). The first claim turned out to be largely true. The second was true but smaller than advertised. The third did not hold up.
The CATIE trial
The NIMH-funded Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) randomised about 1,500 patients with chronic schizophrenia to one of four atypicals (olanzapine, quetiapine, risperidone, ziprasidone) or one typical (perphenazine, chosen because it was widely used and not the most potent first-generation drug). The primary outcome was time-to-discontinuation — a real-world measure of how long patients stayed on each medication. Results, published in NEJM in 2005, surprised many:
- Olanzapine had the longest time-to-discontinuation
- Perphenazine — the typical antipsychotic — performed comparably to most atypicals
- Atypicals did not meaningfully outperform perphenazine on cognitive function or quality of life
- Metabolic side effects were significantly worse on the atypicals (especially olanzapine)
The CATIE summary remains available at nimh.nih.gov.
The CUtLASS trial
The UK CUtLASS-1 trial randomised patients to a typical or atypical antipsychotic (clinician's choice within class) and followed them for a year. Like CATIE, it found no meaningful quality-of-life advantage for atypicals over typicals — and trends actually slightly favoured the typicals. CUtLASS-2 compared clozapine to other atypicals in treatment-resistant patients and confirmed clozapine's superiority.
The EUFEST trial
The European First Episode Schizophrenia Trial (EUFEST) compared low-dose haloperidol to several atypicals in first-episode patients. All groups showed similar symptom improvement; the typical group had higher discontinuation rates due to side effects. This was one piece of evidence supporting current first-episode practice, which generally favours atypicals — but the underlying efficacy was similar.
Side effects: where the classes really differ
Movement disorders (EPS, tardive dyskinesia)
Typicals — especially high-potency ones like haloperidol — cause significantly more acute EPS (parkinsonism, dystonia, akathisia) and have higher long-term risk of tardive dyskinesia. Atypicals have lower rates but not zero rates. See our EPS guide and tardive dyskinesia guide.
Metabolic effects
Atypicals — especially olanzapine, clozapine, and quetiapine — cause significantly more weight gain, insulin resistance, dyslipidaemia, and diabetes risk. Typicals are largely metabolically neutral by comparison.
Prolactin
Both classes can raise prolactin, but high-potency typicals (haloperidol, fluphenazine) and risperidone/paliperidone do so most.
Sedation, anticholinergic effects
Vary widely within both classes — chlorpromazine is heavily sedating, haloperidol is not; olanzapine is sedating, aripiprazole is not.
Cost
Most typicals are dirt cheap as generics. Most atypicals are also generic now (with exceptions like brexpiprazole and lumateperone). The cost gap that drove some 2000s prescribing patterns has largely closed for most agents.
Who might do well on a typical
- Patient who has done well on one historically
- Patient at very high cardiovascular/metabolic risk where avoiding weight gain matters more than minimising EPS
- Cost-constrained settings worldwide where generics dominate
- Patient needing a long-acting injectable from the older line (haloperidol decanoate, fluphenazine decanoate)
Who might do better on an atypical
- First-episode patients (EUFEST and modern guidelines lean atypical)
- Prior poor tolerance of EPS
- Need for mood-stabilising or antidepressant effect (e.g., quetiapine for bipolar depression)
- Treatment-resistant schizophrenia (clozapine is in this class)
Within both groups, individual drugs differ from each other more than the class averages suggest. Choosing between haloperidol and olanzapine is a different conversation from choosing between aripiprazole and lurasidone. Always think drug-by-drug, not class-by-class.
This article is for educational purposes only and is not medical advice, diagnosis, or treatment. Information is summarised from publicly available FDA labelling and peer-reviewed literature. Always consult your prescribing clinician before starting, stopping, or changing any medication. If you or someone you know is in crisis, call or text 988 in the US, or your local emergency number.